Dear HN,
I'm currently lacking ideas for things to do related to ML, I have some compute (a RTX 3070), if the project requires a lot of compute, I may be able to borrow a cluster. Personally, I'd hope it'd require reading quite a few papers and even possibly deploying them, as such, I'd expect each project to take at least a few weeks long.<p>I'd rather not have course suggestions but actual papers/ideas to implement, furthermore, the field isn't of that much importance, most of my experience is in CV / recurrence models but implementing projects in different fields seems equally as interesting!<p>Thanks!
Training and deploying a mini dalle might be suitable depending on your experience, it'll likely require reading a few papers and you can expect it to take at least a few weeks. Not sure if you have the compute for it but you mention being able to get a cluster which would likely be suitable.
recently I liked this researches(and have some more) and would love to work on something like this<p>1. <a href="https://github.com/EricGuo5513/momask-codes">https://github.com/EricGuo5513/momask-codes</a><p>2. <a href="https://github.com/manuelladron/semantic_based_painting">https://github.com/manuelladron/semantic_based_painting</a> (personal fav)<p>I'm also looking for some good projects(and a code buddy), if you are interested in collaborating, ping me.